Most of Emergency lights in our houses fail to work at the right time. Because their Batteries have not exercise charge/discharge for long time, they are not able to take over or in luck cases, they start but end after a short time.

An idea came to my mind to make a SIMPLE timer circuit that will auto-exercise them by turning their power Off / On frequently.

The question is: What is the maximum time can I get of a timer ? even if it is not accurate. One week will be very nice. But is that possible ?

__________________
Its what your friend in Your mind, what you in your friends mind

 

Easily possible, use a PIC to do it - you can then write software to give pretty well any time you like.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Or a simple CMOS oscillator counter like the CD4060 - very useful if you don't want to program and ICs.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez
And http://www.silicontronics.com, same screen name as here.

 

Frankly, Ive got tired of the short life span of most rechargables and have gone to dynamo powered torches. A few cranks keeps them with some charge and when the power does go off a few more keeps them going for 1/2 an hour by which time I can start the genset.

 

Thanks Nigel, but for the cost ($0.4) and availability, I'll go for Hero999 suggestion, cheaper and easy to build. I got its circuit

http://www.coolcircuit.com/circuit/timer_4060/timer.GIF

from:
http://www.coolcircuit.com/circuit/timer_4060/

even this will not fully solve my problem since the maximum time is 320 Minute (5 hours 20 minute).

It says to get more time cascade another Ic. My question is: it says that

T = 2.2 x R2 x C2


Then why don't we simply play with R2 and C2 values?

__________________
Its what your friend in Your mind, what you in your friends mind

 

I am looking for such device in local market with no hope. I am living in a small island where if anything not available in the market, you have to use EXPRESS mail, that doubles the cost and makes it worthless ( around $40 )

__________________
Its what your friend in Your mind, what you in your friends mind

 

I found another circuit promise to make 20 Hours laps, but with this cost will exced PIC price . the circuit is attached and this is the site link:

http://www.uoguelph.ca/~antoon/gadge...ys/relays.html

Thanks

Attached Images

20 hours Timer.JPG (39.0 KB, 15 views)

__________________
Its what your friend in Your mind, what you in your friends mind

 

I think we can acheive long delays of several hours by using a low frequency oscillator and a binary counter.

may be we can use one stage of a schmitt trigger inverter as a square wave oscillator to produce very low frequency...

then we can use a 12 stage binary counter likke the CD4040, where the frequency will divided by 2 at each successive stage... i believe we can get a good delay time before the final stage of CD4040 switches to the high state....

 

Yes, play with the RC values, for example increasing C to 10µF would doulble the delay.

You could also use the CD4521 which has a 24-bit counter.

Also checkout the CD4536 and CD4541 which are designed for this.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez
And http://www.silicontronics.com, same screen name as here.

 

I found CD4541 @ Futurlec.com for $0.45 i'll order it and play with RC circuit to get maximum dealy possible.

Thanks for all

__________________
Its what your friend in Your mind, what you in your friends mind

 

You still haven't said what delay you require.

If you set the oscillator frequency to 1Hz on a CD4541 and used all of the stages, it would give you a delay of 2^24 seconds or 194 days, 4 hours, 20 miniutes and 16 seconds.

I don't know what the minimum frequency the of the oscillator is but I suppose its possible to get delays of over ten years if you really want.

The only problem is capacitors aren't very accurate, 10% at best, so a calculated ten year delay might be anywhere inbetween 9 and 11 years.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez
And http://www.silicontronics.com, same screen name as here.

 

The oscillator capacitor in a CD4060 and CD4541 is supposed to be non-polar. The article uses a polarized 4.7uF electrolytic capacitor which is wrong.

__________________
Uncle $crooge

 

I don't know, I didn't check the datasheet.

Anyway, you can get non-polarised electrolyics with values up to 470µF. The problem with any RC oscillator is the leakage current in the capacitor which might make electrolytics problematic.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez
And http://www.silicontronics.com, same screen name as here.

 

Two polarized electrolytic capacitors connected in series and back-to-back make a non-polar capacitor half the value.

__________________
Uncle $crooge

 

not exact period but if i can make it ON once every week say for half an hour, this will make the Batteris in a better case.

__________________
Its what your friend in Your mind, what you in your friends mind